| 摘要 |
[Problem] Situations involving intelligent machines such as conventional automatic machines or robots are handled according to procedures that have been programmed in advance. Said programs are designed by humans, and must be inputted to a computer installed in the machines, resulting in disadvantages such as requiring significant development time etc. [Solution] This thinking pattern generator converts image information, speech information and language into patterns. The image information is detected by an image detector and converted into patterns corresponding to a target. The speech information is detected by a speech detector and converted into patterns corresponding to speech. The language is converted into patterns composed of phrases and words. The patterns generated from images, speech and language are recorded in a pattern recorder. Relationships between the patterns recorded in the pattern recorder and other patterns are established. The thinking pattern generator learns the thinking patterns of humans, and for certain situations, generates thinking patterns that are similar to a series of thinking patterns that a human would conceive. In addition, by autonomously expanding knowledge by evaluating and recording values for inputted information, it is possible to respond to certain situations in a similar manner to a human. |
